---
title: Zero-Shot and Few-Shot Learning
url: https://www.emergentmind.com/topics/zero-shot-and-few-shot-learning-scenarios
type: topic
---

# Zero-Shot and Few-Shot Learning

Zero-shot and few-shot learning scenarios address the problem of generalizing to novel classes or tasks with little to no annotated training data. These regimes are essential for extending recognition, classification, and generation systems to long-tail or evolving domains where exhaustive labeling is intractable. Both approaches have catalyzed methodological innovation across metric learning, generative modeling, meta-learning, and vision–language foundations, and have recently converged toward unified frameworks for open-domain adaptation.

## 1. Definitions, Scope, and Problem Formalization

Zero-shot learning (ZSL) refers to the setting where a model is expected to recognize or describe instances from classes that are entirely absent from the training set. To enable inference, ZSL methods bridge seen and unseen classes via shared semantic structures—commonly attributes, word vectors, or textual descriptions—so that knowledge can be transferred by associating new inputs to their semantic representations. The canonical ZSL formulation assumes a disjoint partitioning of class labels into seen (training) set $\mathcal{Y}_{0}$ and unseen (test) set $\mathcal{Y}_{1}$, with a semantic embedding function $\pi: \mathcal{Y}_{0} \cup \mathcal{Y}_{1} \to \mathcal{P} \subset \mathbb{R}^{M}$. Classification is achieved by matching an input $x$ to the prototype $\pi(y)$ of each label $y$ based on a learned or fixed compatibility function, often parameterizing $f(x, y) = x^\top W \pi(y)$ [2203.15158].

Few-shot learning (FSL) extends this paradigm by providing a small support set $\mathcal{S}_k$ (often $k=1,5$) of labeled examples for each novel class at test time. The learning goal is rapid adaptation—a model must leverage experience on related base classes to quickly generalize to these novel low-data classes.

Both ZSL and FSL have broad deployment in vision [1703.05175, 1711.06025, 1908.11503], language, tabular [2508.09263], and multimodal evaluation [2509.18839], and in inductive, transductive, and generalized (GZSL/GFSL) regimes.

## 2. Key Methodological Frameworks

### Prototype- and Metric-Based Models

Prototypical networks [1703.05175] provide a foundational FSL approach, learning an embedding $f_\phi$ such that each class is represented as a mean ("prototype"),
$$
c_k = \frac{1}{|S_k|} \sum_{(x_i, y_i) \in S_k} f_\phi(x_i)
$$
Query points are classified via a softmax over (typically squared Euclidean) distances to all prototypes, with squared Euclidean distance shown to be optimal for Bregman-divergence-based clustering and to outperform cosine similarity. This metric-based classifier permits direct ZSL extension by replacing samples with semantic prototypes $c_k = g_\theta(v_k)$, where $v_k$ is the attribute vector for class $k$.

Relation networks [1711.06025], in contrast, learn a non-linear deep metric via a relation module $g_\psi$ applied to concatenated support and query embeddings, trained end-to-end in episodic (meta-learning) regimes to regress relation scores to target equivalence.

Class Adapting Principal Directions (CAPD) [1706.08653] generalize metric transfer by learning per-class projection matrices $W_s$ mapping image features into a semantic space. For zero-shot classes, unseen prototypes are synthesized as linear combinations of seen-class CAPDs, weighted by semantic similarity, and further updated when few-shot samples are available. This framework explicitly supports automatic selection of relevant base classes for each novel class and bias correction in GZSL.

### Generative and Semantic Embedding Approaches

Generative ZSL methods model each class as a distribution in feature space, with parameters (means, covariances) predicted from class-level attributes via learned mappings [1801.09086]. In inductive settings, these mappings are fit to seen classes and extrapolated to unseen ones; in transductive regimes, unlabeled data from unseen classes can further refine these parameters, e.g., via EM updates on Gaussian mixture models.

Recent frameworks integrate generative adversarial networks (GANs) and variational autoencoders (VAEs) both for direct feature synthesis (CVAE, f-CLSWGAN, etc.) and for meta-learned generation [2011.07279, 2102.04379]. For example, Z2FSL [2102.04379] reduces ZSL to FSL by generating synthetic features for unseen classes using a conditional generative model and directly applying few-shot learners such as prototypical networks to these support sets; the entire system is trained end-to-end, allowing FSL loss gradients to influence the generative model, thus improving the utility of synthesized data.

Knowledge-sharing methods [2102.13326] further enrich semantic representations by augmenting textual descriptors with information aggregated from semantically similar classes; GAN-based synthesis on these enriched vectors addresses semantic incompleteness and domain drift.

Semi-supervised scenarios [2308.14119] introduce a KL-divergence penalty into the loss to align empirical prediction frequencies with known class distributions, thereby calibrating confidence between seen (labeled) and unseen (zero-shot) classes even when only a subset of classes are labeled.

### Unified and Multimodal Systems

Unified approaches such as BinBin [2403.03863] recast multi-class tasks as instruction-following binary decisions, combining indirect supervision (pretraining on diverse instruction tasks) and weak supervision (using LLMs to generate labeled examples for zero-shot classes). Such binary-inference transformation enables handling mixed frequent-/few-/zero-shot ("X-shot") label scenarios without architectural changes.

In the multimodal domain, vision–language models (CLIP, SigLIP, multimodal LLMs) have demonstrated strong zero- and few-shot performance in both curated and heterogeneous datasets [2509.18839]. Prompt enrichment and in-context exemplars may further improve zero-shot accuracy, though few-shot gains are dataset- and prompt-sensitive. Prompt-enhanced aggregation, as in PEVA-Net [2404.19168], leverages text-guided fusion to produce strong descriptors in 3D shape recognition; self-distillation from zero-shot to few-shot descriptors robustly boosts accuracy.

For tabular data, ProtoLLM [2508.09263] demonstrates that LLMs can be used to generate feature values for prototype construction via example-free prompts, sidestepping data leakage and token-length constraints, and supporting rapid adaptation in zero-/few-shot settings by fusing LLM priors with empirical support values.

### Data-Augmentation and Optimization-Based Meta-Learning

A detailed taxonomy [2007.15484] summarizes methods for few- and zero-shot learning:
- Data augmentation (e.g., feature hallucination, set operations) enlarges the effective support set.
- Embedding/metric methods (as above) facilitate clustering and comparison.
- Optimization/meta-learning methods (MAML, LSTM meta-learners) rapidly adapt model parameters to new classes.
- Semantics-based techniques incorporate hierarchical or attribute information, variational approaches, and knowledge transfer networks.

## 3. Design Principles and Training Strategies

A recurring principle is episodic/meta-training: models are trained on a sequence of tasks mimicking few-/zero-shot test conditions, enforcing task-adaptive embeddings and calibrating metrics or generative processes for generalization [1703.05175, 1711.06025]. Matching training and test "shot" and "way" improves generalization. For generalized settings (GZSL/GFSL), explicit mechanisms to balance class priors or regularize seen/unseen representation gaps are essential to mitigate seen-class bias [1706.08653].

Semantic regularization strategies such as dual-purpose semantic regularization (DPSR) [2507.01026] enable classifiers to leverage semantic similarities between classes, reducing overfitting to sparse synthetic prototypes and improving decision boundary robustness.

## 4. Practical Performance and Benchmarking

Benchmark comparisons on datasets such as CUB, AwA, SUN, Omniglot, MiniImageNet, CIFAR-100, and ModelNet40 exhibit distinct trends:
- On fine-grained benchmarks (e.g., CUB birds, ICONCLASS iconography [2509.18839]), multimodal LLMs (Gemini-2.5 Pro, GPT-4o) may outperform both vision-only and conventional contrastive VLMs under zero-shot configurations, especially with prompt enrichment.
- In action recognition, generative models achieve substantial accuracy gains in both inductive (no unseen data) and transductive (unlabeled unseen data) settings, and few labeled samples rapidly bridge the performance gap [1801.09086].
- Meta-classifier ensembles (e.g., majority voting, DNNs) can produce modest improvements over single ZSL baselines, but no "overall winner" emerges across datasets or metrics [2203.15158]. Performance remains bounded, with top-1 accuracies frequently below 70% even for leading algorithms.
- On tabular tasks, training-free LLM-based prototype construction matches or surpasses traditional models, particularly in low-shot regimes, with substantial advantages in scalability and privacy [2508.09263].
- Real-world open-domain applications require handling label frequency variability and dataset heterogeneity; unified, instruction-following approaches [2403.03863], and robust prompt engineering are pivotal.

The table below summarizes typical characteristics and strategies in representative scenarios:

| Scenario          | Required Labeled Data    | Key Mechanism           |
|-------------------|-------------------------|-------------------------|
| Zero-shot         | 0 per class             | Semantic transfer, attr/GAN/LLM |
| Few-shot (k-shot) | k per class ($k\ll$ large) | Metric, meta-learning, prototypes |
| Generalized ZSL   | mixed seen/unseen       | Bias correction, feature synthesis, balancing losses |
| Unified X-shot    | 0 to many per class     | Instruction-based binary inference, indirect/weak supervision [2403.03863] |

## 5. Major Challenges and Future Directions

Persistent challenges include:
- Semantic gap and domain shift: Misalignment between visual features and semantic embeddings leads to prediction errors, especially in settings with high intra-class variability or out-of-distribution context [2102.13326, 2406.16143].
- Calibration and bias: Class imbalance (seen/unseen, synthetic/real) necessitates explicit regularization or balancing techniques (e.g., DPSR, KL-penalty [2308.14119]) and careful classifier design.
- Scalability and privacy: Large-scale feature synthesis is computationally costly and may violate ZSL assumptions of "no examples"; strategies that reduce synthetic data (e.g., MSAS in FSIGenZ [2507.01026], privacy-aware mixing in DPS-MOZO#1 [2501.19287]) balance efficiency and robustness.
- Prompt sensitivity: In multimodal/LLM models, prompt quality and alignment with metadata or class descriptions are decisive for zero-shot performance [2509.18839].

Areas identified for further research include: automated prompt optimization, more effective domain-specific adaptations (particularly for medical [2406.16143] or cultural [2509.18839] data), development challenges (e.g., hubness, instability in generative models), and meta-ensemble strategies for robust inference in realistic open-world tasks [2203.15158, 2507.01026].

## 6. Application Domains and Impact

Zero-shot and few-shot approaches have been validated in a spectrum of domains:
- Computer vision: object detection/classification [1706.08653, 1703.05175], 3D shape recognition [2404.19168], fine-grained iconography [2509.18839], and medical imaging [2406.16143].
- Natural language processing: relation extraction and event detection in "X-shot" settings [2403.03863], instruction following, and tabular prediction [2508.09263].
- Semi-supervised and open-world scenarios: leveraging uncurated data (inductive ZSL, ZSL with auxiliary data [2107.04952], or limited supervision [2308.14119]) to close the gap between experimental and practical deployment.

A plausible implication is that the continued synthesis of generative, metric, and instruction-following architectures—combined with more principled prompt engineering and explicit semantic regularization—will remain central in scaling zero-shot and few-shot learning to broader, real-world tasks. The measured progress toward efficient, data-frugal, and robust systems underscores the growing maturity of the field, while simultaneously revealing the practical trade-offs and research questions that persist across domains.

Source: https://www.emergentmind.com/topics/zero-shot-and-few-shot-learning-scenarios